An image processing apparatus that is capable of improving subject detection accuracy with respect to image signals is disclosed. The image processing apparatus applies subject detection processing to an image by using a learning model generated based on machine learning. The image processing apparatus selects the learning model from a plurality of learning models stored in advance, in accordance with characteristics of the image to which the subject detection processing is to be applied.
Legal claims defining the scope of protection, as filed with the USPTO.
1. An image processing apparatus comprising: one or more processors that execute a program stored in a memory and thus function as: a subject detection unit configured to apply subject detection processing to an image by using a learning model generated based on machine learning; and a selection unit configured to select, from learning models that are stored in a storage device for storing a plurality of learning models for use in the subject detection processing, a learning model to be used by the subject detection unit in accordance with characteristics of the image to which the subject detection processing is to be applied, wherein the selection unit selects a first learning model when applying the subject detection processing to an image generated by a first image sensor, the first learning model being acquired by performing machine learning using images corresponding to the first image sensor, the selection unit selects a second learning model when applying the subject detection processing to an image generated by a second image sensor, the second learning model being acquired by performing machine learning using images corresponding to the second image sensor, and the first image sensor and the second image sensor are provided in a same image capture apparatus.
2. The image processing apparatus according to claim 1 , wherein the first learning model is a learning model acquired by performing machine learning using images generated by the first image sensor, and the second learning model is a learning model acquired by performing machine learning using images generated by the second image sensor.
3. The image processing apparatus according to claim 1 , wherein the first learning model is a learning model acquired by performing machine learning using images generated by the first type of image sensor, and the second learning model is a learning model acquired by performing machine learning using images acquired by correcting the images generated by the first image sensor.
4. The image processing apparatus according to claim 1 , the one or more processors further function as a communication unit configured to acquire the learning model to be used by the subject detection unit from the storage device via a network.
5. The image processing apparatus according to claim 1 , wherein the machine learning uses a convolutional neural network (CNN).
6. An image processing apparatus comprising: one or more processors that execute a program stored in a memory and thus function as: a subject detection unit configured to apply subject detection processing to an image by using a learning model generated based on machine learning; and a selection unit configured to select, from learning models that are stored in a storage device for storing a plurality of learning models for use in the subject detection processing, a learning model to be used by the subject detection unit in accordance with characteristics of the image to which the subject detection processing is to be applied, wherein the selection unit selects a first learning model when applying the subject detection processing to an image shot by using a first optical system, the first learning model being acquired by performing machine learning using images corresponding to the first optical system, the selection unit selects a second learning model when applying the subject detection processing to an image shot by using a second optical system, the second learning model being acquired by performing machine learning using images corresponding to the second optical system, and the first optical system and the second optical system are used in a same image capture apparatus.
7. The image processing apparatus according to claim 6 , wherein the first learning model is a learning model acquired by performing machine learning using images shot by using the first optical system, and the second learning model is a learning model acquired by performing machine learning using images shot by using the second optical system.
8. The image processing apparatus according to claim 6 , wherein the first learning model is a learning model acquired by performing machine learning using images shot by using the first optical system, and the second learning model is a learning model acquired by performing machine learning using images acquired by correcting the image shot by using the first optical system.
9. An image capture apparatus comprising an image processing apparatus that comprises: one or more processors that execute a program stored in a memory and thus function as: a subject detection unit configured to apply subject detection processing to an image by using a learning model generated based on machine learning; and a selection unit configured to select, from learning models that are stored in a storage device for storing a plurality of learning models for use in the subject detection processing, a learning model to be used by the subject detection unit in accordance with characteristics of the image to which the subject detection processing is to be applied, wherein the selection unit selects a first learning model when applying the subject detection processing to an image generated in a moving image shooting mode, the first learning model being acquired by performing machine learning using images corresponding to the moving image shooting mode, and the selection unit selects a second learning model when applying the subject detection processing to an image generated in a still image shooting mode, the second learning model being acquired by performing machine learning using images corresponding to the still image shooting mode.
10. The image capture apparatus according to claim 9 , further comprising a first image sensor and a second image sensor, wherein in the still image shooting mode, the first image sensor is used and the second image sensor is not used, and in the moving image shooting mode, the second image sensor is used and the first image sensor is not used.
11. The image capture apparatus according to claim 10 , wherein in the still image shooting mode, an optical finder is in use, and in the moving image shooting mode, the optical finder is not used.
12. The image capture apparatus according to claim 9 , wherein the first image sensor is used for acquiring an image for exposure control in the still image shooting mode.
13. An image processing method executed by an image processing apparatus, comprising: applying subject detection processing to an image by using a learning model generated based on machine learning; and selecting, in accordance with characteristics of the image to which the subject detection processing is to be applied, a learning model to be used in the subject detection processing to be applied to the image from a storage device for storing a plurality of learning models for use in the subject detection processing, wherein in the selecting, a first learning model is selected when the subject detection processing is applied to an image generated by a first image sensor, the first learning model being acquired by performing machine learning using images corresponding to the first image sensor, in the selecting, a second learning model is selected when the subject detection processing is applied to an image generated by a second image sensor, the second learning model being acquired by performing machine learning using images corresponding to the second image sensor, and the first image sensor and the second image sensor are provided in a same image capture apparatus.
14. A non-transitory computer-readable medium storing thereon a program for causing a computer to function as an image processing apparatus comprising: a subject detection unit configured to apply subject detection processing to an image by using a learning model generated based on machine learning; and a selection unit configured to select, from learning models that are stored in a storage device for storing a plurality of learning models for use in the subject detection processing, a learning model to be used by the subject detection unit in accordance with characteristics of the image to which the subject detection processing is to be applied, wherein the selection unit selects a first learning model when applying the subject detection processing to an image generated by a first image sensor, the first learning model being acquired by performing machine learning using images corresponding to the first image sensor, the selection unit selects a second learning model when applying the subject detection processing to an image generated by a second image sensor, the second learning model being acquired by performing machine learning using images corresponding to the second image sensor, and the first image sensor and the second image sensor are provided in a same image capture apparatus.
15. A non-transitory computer-readable medium storing thereon a program for causing a computer included in an image capture apparatus to function as an image processing apparatus that is included in the image capture apparatus and comprises: a subject detection unit configured to apply subject detection processing to an image by using a learning model generated based on machine learning; and a selection unit configured to select, from learning models that are stored in a storage device for storing a plurality of learning models for use in the subject detection processing, a learning model to be used by the subject detection unit in accordance with characteristics of the image to which the subject detection processing is to be applied, wherein the selection unit selects a first learning model when applying the subject detection processing to an image generated in a moving image shooting mode, the first learning model being acquired by performing machine learning using images corresponding to the moving image shooting mode, and the selection unit selects a second learning model when applying the subject detection processing to an image generated in a still image shooting mode, the second learning model being acquired by performing machine learning using images corresponding to the still image shooting mode.
16. A method executed by an image processing apparatus provided in an image capture apparatus comprising: applying subject detection processing to an image by using a learning model generated based on machine learning; and selecting, from learning models that are stored in a storage device for storing a plurality of learning models for use in the subject detection processing, a learning model to be used by the subject detection unit in accordance with characteristics of the image to which the subject detection processing is to be applied, wherein in the selecting, a first learning model is selected when the subject detection processing is applied to an image generated in a moving image shooting mode, the first learning model being acquired by performing machine learning using images corresponding to the moving image shooting mode, and in the selecting, a second learning model is selected when the subject detection processing is applied to an image generated in a still image shooting mode, the second learning model being acquired by performing machine learning using images corresponding to the still image shooting mode.
17. An image processing method executed by an image processing apparatus, comprising: applying subject detection processing to an image by using a learning model generated based on machine learning; and selecting, from learning models that are stored in a storage device for storing a plurality of learning models for use in the subject detection processing, a learning model to be used by the subject detection unit in accordance with characteristics of the image to which the subject detection processing is to be applied, wherein in the selecting, a first learning model is selected when the subject detection processing is applied to an image shot by using a first optical system, the first learning model being acquired by performing machine learning using images corresponding to the first optical system, in the selecting, a second learning model is selected when the subject detection processing is applied to an image shot by using a second optical system, the second learning model being acquired by performing machine learning using images corresponding to the second optical system, and the first optical system and the second optical system are used in a same image capture apparatus.
18. A non-transitory computer-readable medium storing thereon a program for causing a computer to function as an image processing apparatus comprising: a subject detection unit configured to apply subject detection processing to an image by using a learning model generated based on machine learning; and a selection unit configured to select, from learning models that are stored in a storage device for storing a plurality of learning models for use in the subject detection processing, a learning model to be used by the subject detection unit in accordance with characteristics of the image to which the subject detection processing is to be applied, wherein the selection unit selects a first learning model when applying the subject detection processing to an image shot by using a first optical system, the first learning model being acquired by performing machine learning using images corresponding to the first optical system, the selection unit selects a second learning model when applying the subject detection processing to an image shot by using a second optical system, the second learning model being acquired by performing machine learning using images corresponding to the second optical system, and the first optical system and the second optical system are used in a same image capture apparatus.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
April 3, 2019
January 19, 2021
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.